Joe Schmoe
Posts: 4277
Joined: Sun Jan 15, 2012 1:11 pm

Re: the teething troubles... (not the launch)

Thu May 17, 2012 5:50 pm

adlambert wrote: I've been using an MS branded Mouse, an optical wheel mouse. It's been absolutely fine.
My keyboard is a very cheap PCLine one, that's fine too. Never had any problems when using them together.
Now the Pi is only accessed using Putty from a Laptop over SSH so no keyboard or mouse needed. When in command line mode it is in its element and it's a great way to learn.
Totally agree with your last sentence!

And, just to clarify, when I referred to "luxuriously designed", by "luxurious", I meant "designed to spec". Most of the crap people buy today is *not* to spec. And that's the problem.

Further, when I first heard about the Pi, I, like many others, I am sure, assumed that any old charger laying around would work. Now, we all know differently...
And some folks need to stop being fanboys and see the forest behind the trees.

(One of the best lines I've seen on this board lately)

robwriter
Posts: 114
Joined: Sat Feb 25, 2012 6:26 pm

Re: the teething troubles... (not the launch)

Thu May 17, 2012 6:00 pm

I doubt anyone needs my opinion, but for the Pi to succeed it's going to need to be able to do the simple stuff, like web browsing, at a reasonable speed. When I code I rely on Google. Partially because I'm rubbish, but mostly because it's just easier that way. If I need to go to my PC to look at websites to help me code python then I may as well use the PC to code.

No one who's sensible is expecting a super fast piece of kit, but when I remember that my first web browsing experience was on a 486 SX33 with 4Mb RAM, I find it hard to fathom the Pi can't be made to browse the web at a decent speed.

I'm sure the USB problem is on the foundations radar - it's being over exaggerated by many people, but it is an issue that'll need to be fixed. At the very least we need to work out what works and doesn't and why so people can consistently get a setup that works - right now it seems like a lucky dip.

No one really has the right to complain though, as we all bought (or are trying to buy) boards knowing we are the testers, so we should really channel this energy to making things better rather than posting on forums. Me included.

User avatar
AndrewS
Posts: 3625
Joined: Sun Apr 22, 2012 4:50 pm
Location: Cambridge, UK
Contact: Website

Re: the teething troubles... (not the launch)

Thu May 17, 2012 6:05 pm

robwriter wrote:No one who's sensible is expecting a super fast piece of kit, but when I remember that my first web browsing experience was on a 486 SX33 with 4Mb RAM, I find it hard to fathom the Pi can't be made to browse the web at a decent speed.
Remember watching JPEG images being drawn a single line at a time? ;)
If you want speedy browsing on the Pi, you could always use the lynx text-mode browser 8-)

MarshallBanana
Posts: 24
Joined: Sat Jan 21, 2012 9:19 pm

Re: the teething troubles... (not the launch)

Thu May 17, 2012 7:18 pm

AndrewS wrote:If you want speedy browsing on the Pi, you could always use the lynx text-mode browser 8-)
no joke, you can start the links-textbrowers in "graphical" mode (-g switch) so images are included... works on framebuffer-device and "windowed" in X... i often use this on my netbook when i don't need/want flash-bloatware and fancy javascript-magic....

User avatar
Jim Manley
Posts: 1600
Joined: Thu Feb 23, 2012 8:41 pm
Location: SillyCon Valley, California, and Powell, Wyoming, USA, plus The Universe
Contact: Website

Re: the teething troubles... (not the launch)

Thu May 17, 2012 8:04 pm

Having been working on a Pi remotely for over a month, and with one in front of me for five whole days (time flies when you're having fun! :) ), I'm of the firm opinion that it's perfectly fine for its intended use and when properly equipped with adequate peripherals. If you tried to run a top-end multi-core CISC system with multiple GBs of RAM, multi-TB hard disks, a top-end GPU card, etc., using a 300 watt power supply, guess how well that system would boot up, much less handle a keyboard and mouse? You'd be lucky if you even got the Power On Self Test (POST) beeps to be audible. If people with boards post to a list on the wiki the power adapters that work and don't work by manufacturer, model number, specs as stated on the device, etc., those following behind in receiving their boards will know what works and what to avoid. It won't be too long before a reasonably-priced, sufficiently-capable choice of devices is identified.

If you don't happen to have a minimal keyboard and mouse lying around, you can easily get as many as you want from your local recycling center - businesses toss them by the millions every year as they upgrade and USB has been around long enough that several generations of machines have had them, by now. Heck, pick up as many as you can carry and test to see how many work! Even an old PS/2 keyboard or mouse can be made to work using a plug adapter that costs less than a plain cup of coffee these days.

If you're going to plug more than a minimal mouse and keyboard into the USB ports on the board, you're going to need a powered USB hub - period. The intended educational use really doesn't require more than a keyboard - running a GUI with a mouse is a bonus at the very least. For those kvetching about how slow web pages load, I would invite you to click on the Show Page Source option on a particularly slow-loading page and compare it to that paragon of spartan bit frugality, the Google home page (with the plain logo, not a special edition commemorative page). Comparing the performance of any computer today trying to load and run all of the typical crapware sent in response to an HTTP request with any personal computer processing a typical web page of the 1990s (the rough equivalent of the Pi's CPU) is like comparing the Space Shuttle to the Wright Flyer. The Shuttle is a whole lot more expensive and complicated, but, it will get you going 17,000+ mph and back home (if operated within sensible limitations, not in sub-freezing weather or when icebergs are falling off the external fuel tank - there's a couple of lessons about patience that "grown-ups" need to relearn and pass along to kids) - the Wright Flyer, not so much.

There is a long laundry list of things that need to be worked on, much less finished, before the Pi will be ready for prime time, i.e., any given classroom dedicated to teaching kids computing technology principles. However, like the board itself, it's not going to happen unless a lot of volunteers pitch in and help do the laundry, a lot more than it took to develop the hardware. The nice thing about software is that, if at first you don't succeed, try, try again, and it's much easier to divvy up software into many inter-operating pieces that people can work on more independently than is usually the case with closely-integrated hardware. There is also the strongly-hoped-for potential for students to participate in the work - what better way to learn intellectually-stimulating and economically-useful skills than to improve your own possession? I think the low cost of entry for the Pi is absolutely key to its widespread adoption, along with the fact that it's nearly state-of-the-art technology, and therefore ideal for getting people up-to-speed on modern, low-power, high-performance graphics systems. The GPU in the Pi is reportedly four times more powerful than that in the iPhone 4, which is an ~$800 (unlocked) device that's taking a lot of market share from long-established game system manufacturers. I still can't believe the Pi is less than $100, and being under $50 delivered is nothing short of miraculous.

That's why I'm focused on getting 3-D graphics software running that exploits the entire system to its full potential, initially through the Pi-finity! educational game software I'm developing. However, I think we can go way beyond that to making manipulation of 3-D objects on the Pi as ordinary and de rigueur as handling documents in window-based software is today. There is potential in the Pi for an actual paradigm shift in user abilities, and the current surge in interest in low-cost, high-resolution 3-D object printing is absolutely complementary to what the Pi is capable of in 3-D software. It's too soon for this year's Maker Faire in California, but, by next year's Faire, I hope to see a strong tie-in between the Pi and 3-D object design and production.

In the end, it will be all right. If it's not all right, it's not the end ... ;)
The best things in life aren't things ... but, a Pi comes pretty darned close! :D
"Education is not the filling of a pail, but the lighting of a fire." -- W.B. Yeats
In theory, theory & practice are the same - in practice, they aren't!!!

Eros
Posts: 26
Joined: Tue Feb 28, 2012 5:16 am

Re: the teething troubles... (not the launch)

Thu May 17, 2012 8:24 pm

Sorry for the long post but i think much of this is just operating the device, and understanding the limitations, but not accepting them.

I dont have mine yet, though am confirmed in the next batch and am waiting on one from RS and one from Newark.

I always see things subjectively and it has been pointed out many many times what the device is intended to do, push it too far beyond that and sure the user experience will not be as nice as it is on a cheap $400 PC (or perhaps even cheaper).

One point id like to make is the whole issue about USB devices, this could have been an oversight but from everything iv read it smacks as a power supply issue. Basically, not giving the device enough power will cause you problems. I see this as a user fix personally and not one the foundation is responsible for.

There are many cross references to do with hardware and what it is intended to do that are exactly the same when applied to PCs....
Examples,
RAM - Max out your ram on a motherboard, and then try and overclock it, even the nicest and most expensive motherboard can run into problems, and no you dont have bad ram, you are saturating the bus controller which might work fine with 50% of the ram when massively OCd, but not when it has 100%. Had so many people send sticks of ram back claiming it was dud when the problem really is that they are operating both the ram and the motherboard way way out of spec.

PCs locking up randomly - No not always caused by windows, more often caused by a bad power supply (see the parallel here?) When people complain when ever they put any load on a PC either on the CPU or the GPU, the pc locks up... 90% of the time it is caused by an aged/inadequate psu. What is the first thing people look at? Temperatures... *facepalm* when normally its caused by the PSU dropping voltages when larger current is drawn when under load.

robwriter
Posts: 114
Joined: Sat Feb 25, 2012 6:26 pm

Re: the teething troubles... (not the launch)

Thu May 17, 2012 9:23 pm

AndrewS wrote:
robwriter wrote:No one who's sensible is expecting a super fast piece of kit, but when I remember that my first web browsing experience was on a 486 SX33 with 4Mb RAM, I find it hard to fathom the Pi can't be made to browse the web at a decent speed.
Remember watching JPEG images being drawn a single line at a time? ;)
If you want speedy browsing on the Pi, you could always use the lynx text-mode browser 8-)
I remember JPEG's loading slowly. But I was using a 28.8K modem.

User avatar
scep
Posts: 1062
Joined: Sun Nov 20, 2011 8:53 am

Re: the teething troubles... (not the launch)

Thu May 17, 2012 9:29 pm

Jim Manley wrote:Having been working on a Pi remotely for over a month, and with one in front of me for five whole days (time flies when you're having fun! :) ), I'm of the firm opinion that it's perfectly fine for its intended use and when properly equipped with adequate peripherals
...
In the end, it will be all right. If it's not all right, it's not the end ... ;)
A very comprehensive and rational summing up by someone who knows his stuff. Thanks Jim.

plugwash
Forum Moderator
Forum Moderator
Posts: 3417
Joined: Wed Dec 28, 2011 11:45 pm

Re: the teething troubles... (not the launch)

Fri May 18, 2012 5:30 pm

Eros wrote: One point id like to make is the whole issue about USB devices, this could have been an oversight but from everything iv read it smacks as a power supply issue. Basically, not giving the device enough power will cause you problems. I see this as a user fix personally and not one the foundation is responsible for.
We have been told the Pi was supposed to run off a USB PSU. Reading the specs for those they are supposed to keep the voltage between 4.75V and 5.25V

We have been told that the Pi is only supposed to be able to power low power USB devices. A low power USB device may draw up to 100ma and is required to operate off voltages down to 4.4V.

Users have measured the cold resistance of the polyswitches on the USB ports at 5-7 ohms, given how polyswitches work the hot resistance is likely to be higher than the cold resistance. At 100ma 7 ohms leads to a volt drop of 0.7V that means the main 5V line must be at least 5.1V which given the resistances of the supply cable and the input polyfuse means that the voltage at the PSU needs to be at arround 5.2V.

Afaict the reason most users are getting away with this is that most keyboards and mice don't actually draw anywhere near 100ma and that most better quality USB PSUs are actually towards the upper end of the allowed range.

Personally If I have power issues the first thing I would do is either solder a wire across those polyswitches or remove them and replace them with zero ohm links (personally I would do the latter but I have a set of heated tweezers which will pop them off in seconds which is a luxury most people don't have)

P.S. Some screwups are to be expected with a project of this complexity and I don't think it reflects badly on the foundation as long as an appropriate fix is in for the next release.

User avatar
mahjongg
Forum Moderator
Forum Moderator
Posts: 11850
Joined: Sun Mar 11, 2012 12:19 am
Location: South Holland, The Netherlands

Re: the teething troubles... (not the launch)

Sat May 19, 2012 2:19 pm

just one remark here, the USB specs say that at 4.4 Volt, a low speed USB device should be able to keep up USB communication without fail, but that doesn't mean the whole USB connected device should work at 4.4 volt! In fact the specs state that for the whole device the normal 5 V +- 5% rules apply, so the minumum voltage for a low speed USB device, such as a mouse or keyboard, is still 4.75 Volt!

Also, it seems that PSU specs are mistakenly speculating that there won't be any voltage loss in the power chan ever, so they can claim the full 5% tolerance, obviously that is fictional!
A more reasonable tolerance for a PSU would be +5% and -2%, and I suspect that real PSU (not "chargers") would keep to that, at least if they are professionally designed, which unfortunately isn't always the case with the very cheap stuff, but should be the case with branded stuff.

Return to “General discussion”